Tropical Storm Hermine is forecast to strengthen as it approaches landfall near Florida’s Big Bend/ Apalachee Bay.
Andy Horowitz’s recent New York Times opinion piece on the new flood insurance rate maps for New Orleans contains a number of inaccurate statements, fails to address the fundamental issue, and amounts to nothing more than picking on the “cash-strapped New Orleanians” as they continue the long process of recovering from catastrophic levee failures that occurred over a decade ago. Many of the shortcomings of this piece have been pointed out here and here. In addition to the factual errors that these critiques raise, Prof. Horowitz perpetuates a fundamental misunderstanding of flood risk and how our nation manages it. In no way do the new maps suggest that flood risk has been eliminated, as Prof. Horowitz claims. In perpetuating this inaccurate cliche, the article dances around but does not directly address a fundamental flaw in the National Flood Insurance Program.
“Safe from Flooding” is a Fiction
Prof. Horowitz’s statement that the new flood maps “declared most of New Orleans safe from flooding” is incorrect. The new maps merely moved some areas from the 1% annual chance flood risk area, the “100-yr flood zone” or the “special flood hazard”, to areas that are estimated to have an 0.2% chance of flooding during any given year, the so-called “500-yr flood zone.” (A flood that has a 1% annual chance will have an average return of 100 years, while a 0.2% annual chance corresponds to an average return period of 500 years.) This adjustment is an accurate reflection of the risk reduction provided by the Federal, state, and local investment in an improved levee, floodwall, and pump station system. (However, FEMA’s methodology does not account for the flood risk reduction benefits provided by numerous coastal restoration and community resilience projects.)
The article wrongly perpetrates the misconception that properties in the 500-yr floodzone are not considered at risk. As Prof. Ed Link, a nationally recognized flood risk expert at the University of Maryland, put it: “there is no mention of safe within the language of the FEMA program or the maps.” In repeating this myth, Prof. Horowitz completely misses the point that these properties are at risk, that the maps accurately depict this risk, and that these property owners should pay into the flood insurance risk pool.
New Orleans’ Flood Maps Advanced the State of the Practice
Prof. Horowitz is also incorrect to state that the flood risk analysis did not account for sea level rise, subsidence, or the probability of levee failures. All of these factors were considered in the technical analysis. It was a complex and tricky process, and there are legitimate questions about the assumptions and methods used. But what he bemoans as “six years of waiting and wrangling” was in fact a series of intensive consultations between national and local technical experts dealing with the very challenges that Prof. Horowitz claims were ignored. This is what it takes develop, in an open and public manner, flood risk maps that adhere to national standards while also including local data and knowledge. The result is likely the nation’s most comprehensive, computationally intensive, mathematically rigorous, and transparent attempt to estimate flood risk for a region. The northeast coast in the aftermath of Hurricane Sandy is the only region that we know of that comes close to the effort to create flood risk maps that accounts for changes in the climate and landscape.
In the face of these challenges, this is the effort it takes to have hydrologically correct flood risk maps for our coastal cities. Further, hydrologically correct flood risk maps is a necessary condition for an actuarially correct flood insurance program, the stated goal of many years of effort at flood insurance reform.
The implicit assumption throughout the article is that the maps are an inaccurate reflection of flood risk for the region, but the article does not attempt to make that case. In fact, his statement that the maps “reflect the most optimistic possible assessment” is a tacit acknowledgement that the flood risk depicted on the maps is in fact within the confidence interval of the technical analysis. It is clear the Prof. Horowitz does not like the outcome of the maps, and he does raise legitimate concerns regarding the outcome. However, in attacking the maps and the process that created the maps, he failed to address the fundamental problem with flood insurance in New Orleans or elsewhere in America.
Flood Insurance Risk Pool Needs to Include the 500-yr Floodzone
In moving areas from the 100-yr to 500-yr flood zone, the maps in no way imply that areas are safe. In fact, they explicitly state that there is a 0.2% chance that flooding will occur during any given year. The fundamental problem is that the flood insurance program currently exempts these properties in the 500-yr flood zone from flood insurance requirements. Because they are not required to pay into the risk pool, these residents are less likely to be aware of their flood risk. Expanding the risk pool to include these properties increases flood risk awareness of these residents. It is also another necessary condition for an actuarially correct and financially sustainable flood insurance program. The failure to address this fundamental flaw in the flood insurance program largely explains why attempts to reform the flood insurance program have failed. Arbitrarily changing the flood maps for New Orleans because Prof. Horowitz does not like the outcome does nothing to address this flaw.
Flash flooding across most of Louisiana along with parts of Texas, Arkansas, Mississippi, and Tennessee.
6 fatalities reported and 2 people missing
Over 5,000 people rescued from flood waters including swift water rescues
Estimated 18,000 homes suffer some degree of flood damage, consisting of 11,000 in Louisiana, 5,000 in Texas, and 2,000 in Mississippi.
DisasterMap.net Estimate: $1.2 billion in damage to structures, vehicles, and due to closure of I-10.
This page was last updated on April 19.
A severe weather system driven by an “atmospheric river” created a prolonged rain event that caused deadly and destructive flashing flooding in Texas, Louisiana, Mississippi, Arkansas, and Tennessee. Parts of Louisiana received as much as 20″ of rain. Additionally, prolonged onshore winds caused minor coastal flooding in Texas and Louisiana. 6 possible storm related deaths so far and an estimated 6,000 people have been forced from their homes.
Summary of Hazards:
Heavy rainfall induced flash flooding and coastal flooding.
Observed cumulative (March 4 – 11) rainfall amounts include:
– Nearly 22″ observed near Monroe, La
– Over 12″ observed in Covington, La and points north
– A large region in central Louisiana that experience 10″ or greater the inlcudes Shreveport, Monroe, and Natchitoches
– A major 5″ contour line covering a large region that includes Houston, the outskirts of Dallas, Memphis, and Jackon, Ms
– A second 5″ contour line that includes Baton Rouge, Slidell, Hattisburg, and McComb.
The cumulative precipitation maps shown below were generated by DisasterMap.net using data from the National Weather Service. The 7-day cumulative precipitation map (CONUS) was downloaded for the March 4 – 11, 2016 period as a point shapefile. From the point shapefile, a raster interpolation was completed and then contour lines were generated using the raster. The processing and mapping was completed using QGIS.
The first map below shows the generated contours lines and the original NWS point dataset (at 50% transparency.) The second map below shows the 5″ and 10″ contour lines with the original point dataset (filtered to only show values over 2″ and at 50% transparency) over a OpenStreetMap basemap (served by MapQuest.)
Wind driven high tides extended from southern tip of Texas to the middle of the Florida panhandle. 3-4 ft tides measured in and between Galveston Bay and Bay St Louis. Over 4 ft high tides were observed at two stations along the central Louisiana coastline.
DisasterMap.net’s Estimate of Impacted Populations:
This section presents the results of an initial attempt to estimate size of the population impacted by this event. It is based on geoprocessing steps completed in a Geographic Information System (GIS). It utilizes two hazards layers, rainfall contours and flooded areas, along with 2010 US Census population data. It represents a rough estimate at a particularly moment in time. The data contains uncertainties, and the methods while quick are imperfect. In addition, this event continues to unfold. The results are at best a rough estimate representing the particular moment in time captured by the hazard data.
Using QGIS software, the 2010 US Census Demographic Profile 1 geodatabase was clipped by the 5″, 10″, and 12″ rainfall contours described above. This procedure identified the Census tracts that experienced rainfall by these amounts or greater. For the clipped tracts it was then possible to sum the population and housing units. Naturally, being within any of these contours does not necessarily imply direct flood impacts. However, it does mean that indirect flood impacts, such power outages, isolation by waters, cut off from medical services, or work/employment impacts, are likely. This is particularly true for the 12″ contour.
The table below summarizes the population and houses within each of Census tracts that overlapped the different contour lines, while the map shows these Census tracts color coded by population.
|Within 12″ Contour||Within 10″ Contour||Within 5″ Contour|
|Total Housing Units||203,600||404,204||3,774,001|
Applying a similar procedure using a flooded area polygon obtained from the Dartmouth Flood Observatory, it is possible to obtain an initial estimate of the population and homes located within Census tracts that experienced flooding. For this early assessment, the “MSW_2016066_100W040N_3D3OT_V” flood zone shapefile was obtained from their repository, and then the Census DP1 layer was clipped by it. The Dartmouth Flood Observatory provides continuous and global monitoring of floods using stream gauge and satellite data. Their flooded areas shapefiles are produced through a land/water classification of primary MODIS satellite data, which has a 250 m spatial resolution. Our steps identified Census tracts that overlapped with the flooded areas identified by Dartmouth. Importantly, not everyone within a Census tract that overlaps the flooded areas actually flooded. They may have an elevated home or the flood may be on the other end of their tract. Still this provides a reliable early measure of the magnitude of the emergency response.
The table below summarizes the population and houses within each of Census tracts that overlapped the flooded areas identified by the Dartmouth Flood Observatory:
|Within Census Tracts that Experienced Flooding|
|Population Under 5 yrs Old||129,358|
|Population 85 yrs and Older||34,029|
|Total Housing Units||1,067,395|
|Occupied Housing Units||847,452|
|Owner-Occupied Housing Units||655,997|
|Renter-Occupied Housing Units||191,455|
Flood Data Source:
Citation: Brakenridge, G.R., Anderson, E., Kettner, A.J., Slayback, D. and Policelli, F., March 13, 2016, “Current Flood Conditions, MSW_2016066_100W040N_3D3OT_V “, Dartmouth Flood Observatory, University of Colorado, Boulder, CO, USA, http://floodobservatory.colorado.edu/.
DisasterMap.net’s Estimate of Economic Impacts:
DisasterMap.net estimates at least $1.2 billion in damage to homes, vehicles, and due to the closure of I-10. This includes an estimated $900 million in flood damage to structures.
Flood Damage to Buildings and Contents
An estimated 18,000 structures have suffered flood damage. If the average cost of damage is $50,000 each, then the total direct damage to structures could reach $900 million. Additionally the contents of 18,000 homes also suffered flood damage. If they average $10,000 in contents damage per house, then that would total $180 million
Flood Damage to Vehicles
While flooded vehicles have been documented in media and social media, no estimate of flooded vehicles have been made available. In the absence of a ground based assessment, it is reasonable to assume that at least 6,000 vehicles flooded (1 for every 3 homes that flooded). If they averaged $10,000 in damage, then the total would be $60 million.
Cost of I-10 Closure
DisasterMap.net estimates that the I-10 closure will have an economic impact of the order $22.5 million, based on a 2.5 day closure and $9 million per day.
A Texas Department of Transportation and Development press release states that I-10 was closed on the morning of Wednesday, March 16. At approximately 8 pm on March 18, the interstate reopened to traffic after the water receded and emergency repairs were made to the eastbound lane. The interstate was closed for approximately 60 hrs.
(DisasterMap.net has previously assessed that “According the NWS river gauge nearest I-10, the Sabine river was at 7.36 ft at 9am on 3/16, approximately when the interstate flooded. This gauge is forecast to remain at or above this height for approximately 55 hrs, or just over 2 days.”)
A study of two earlier interstate closures, due to flooding and an avalanche, found that the closures cost on average $9 million per day. For a 4 day closure due to the flood, they found $47 million in economic damages, $11.75 million per day. For the 4 day closure due to the avalanche, they estimated $28 million, $7 per day. Averaged, that gives $9 million per day.
Additional Damage not Estimated
In addition to the classes of damage described above, other types of known flood damage have not been tallied here. This includes crop losses, infrastructure (specifically a number of bridges, culverts, and roads are known to be damaged), and business and employment impacts. Additionally, emergency response and long term recovery costs have not been tallied.
Below is a summary of key impacts reported by the press and government agencies:
Social Media from the Scene:
Stay Informed in Real time with DisasterMap.net:
Our homepage includes a Leaflet webmap with key hazard warning layers, and Full Size Map let’s you interact with the data to find out information specific to your location. It also provides official warnings and guidance from the relevant government agencies through an RSS feed. Our Twitter feed gives you reports, photos, and videos direct from the ground. Follow us on social media to get updates. We strive to bring you the sights and sounds of the impacted areas.
Click the “View Full Size Map” to get to the fully interactive webmap, which include shelter locations. Click one of the blue crosses to get the name, address, and capacity of any shelter that is part of FEMA’s National Shelter System.
Use our address search to view conditions for areas where you are concerned. And, our real time traffic layer displays any roads closed due to flooding.